DIY Audio Projects Forum
https://diyaudioprojects.com/Forum/

Assist with waveform
https://diyaudioprojects.com/Forum/viewtopic.php?f=9&t=7030
Page 1 of 1

Author:  gofar99 [ 08 Jan 2021, 20:30 ]
Post subject:  Assist with waveform

Hi Folks, Another DIYer sent me the attached waveform. It is taken after a bridge of UF4007 rectifiers at about 250 VDC into a 100uf cap. I don't recall ever seeing any quite like this one and would like your thoughts.

Good listening
Bruce
Attachment:
GetAttachmentThumbnail.png

Author:  M. Gregg [ 09 Jan 2021, 07:55 ]
Post subject:  Re: Assist with waveform

Hi,

Did the DIYer send the circuit diagram? Or a picture of the rectifier and transformer?
Just a mad guess is the electrolytic cap connected backwards. if there is no loading where is the charge retention?
The diodes should only conduct when the applied voltage to the diodes is higher than the charge voltage on the smoothing cap.
Obviously the supplied voltage has to also exceed the forward voltage drop on the diodes as well. :)

Just for fun compare;
https://www.google.co.uk/search?q=full+ ... 55&bih=687

I find it hard to believe;
The circuit it correct
or the diodes or cap are not faulty.

The only other thing is a shorted power Tx or wrong connection.
Is it full wave bridge or full wave CT?
Or is it a doubler circuit?

Regards
M. Gregg

Author:  gofar99 [ 09 Jan 2021, 08:44 ]
Post subject:  Re: Assist with waveform

Hi, Much my thoughts as well. But with abour 250VDC the cap would blow in short order. I attached the schematic. The trannie is not shown, but connections are as you would expect. Separate winding and no CT. DIYER is experienced service tech for pro gear (many years). Trannie is rated at well above needed current. I asked him to provide a non-differential spectrum and a dc side waveform. Also to replace the cap. Like I said in the first post, I have never seen a waveform quite like this. Only time similar was one with zeners going on it that truncated the peaks on the rectified DC.

Good listening
Bruce
Attachment:
PS.png

Author:  M. Gregg [ 09 Jan 2021, 08:49 ]
Post subject:  Re: Assist with waveform

Hi,

Ask the diyer to disconnect the regulator from the smoothing cap and send a pick of the ripple DC across the cap.
Obviously the diyer will have to watch for overvolt on the 100uF.
It could be the regulator oscillation maybe?

Then connect the regulator and disconnect the load and send the pick of DC across the 100uF again.
Then connect the load to the reg if its shows the problem load the reg with just a resistor and see if the same thing occurs?

ie reg instability or load instability.

Regards
M. Gregg

Author:  M. Gregg [ 10 Jan 2021, 04:49 ]
Post subject:  Re: Assist with waveform

Hi,

The only other thing I can think of is looking at the output cap C51.
What is the specified value for the regulator type?
The output cap value is sometimes specified for the stability of the regulator chip.

NB try a 0.01uF across the input to the reg as well. :) (not much to remove commutation noise on the DC side of the rectifier).

Regards
M. Gregg

Author:  HT Performance [ 10 Jan 2021, 17:00 ]
Post subject:  Re: Assist with waveform

I´d would suspect the batch of diodes...

Author:  Les [ 10 Jan 2021, 22:36 ]
Post subject:  Re: Assist with waveform

I agree with HT. Seems like rectifiers aren't working. Where did he source the diodes (was it a reputable source (Mouser, DigiKey, etc.) or Alibaba)?
Could there be a short to ground in the windings on the output side of the transformer?

Some things that peak my curiosity are the size of C50 & 51 and the inclusion of D23. Most of the diagrams I've seen either do not have a capacitor on the Adj pin or no larger than 10microf. I've seen no other diagram including a diode in parallel with the voltage divider. Output capacitor (C51) should be 1microf.
I've built plenty of low voltage regulators (based around the 317) and have never included a capacitor on Adj and never a bypass diode on the voltage divider. I've also tended toward larger output capacitors on the output with few problems - you can go too large :blush: .

It's been a while, but the last time I saw anything similar to that waveform, the bridge had shorted and there were cascade failures following. That was a package bridge, not discrete components. I've only ever seen package bridge rectifiers short out.

Note:
I'm not knocking Alibaba in particular, but rather the fact that there are a lot of questionable and blatantly fake products sold through them.

Page 1 of 1 All times are UTC - 6 hours [ DST ]
Powered by phpBB® Forum Software © phpBB Group
http://www.phpbb.com/